- Email: [email protected]
Origin of rate-acceleration in ester hydrolysis with metalloprotease mimics
Bioorganic & Medicinal Chemistry 8 (2000) 647±652
Origin of Rate-Acceleration in Ester Hydrolysis with Metalloprotease Mimics Dong H. Kim* and Soo Suk Lee Center for Biofunctional Molecules and Department of Chemistry, Pohang University of Science and Technology, Pohang 790-784, South Korea Received 23 August 1999; accepted 26 November 1999
AbstractÐMimics of carboxypeptidase A, a prototypical metalloprotease, were synthesized by linking macrocyclicpolyamines to the primary side of b-cyclodextrin followed by complexing with Zn(II). These enzyme mimics exhibit saturation kinetics in hydrolysis of p-nitrophenyl acetate (PNPA) and enhance the rate of hydrolysis reaction by almost 300-fold. The eective molarities (EM) of the mimics range from 0.2 to 1.9 M. Origin of the rate acceleration was examined: the reactivity of Zn(II) complexes of [12]aneN3 [12]aneN4, and [14]aneN4 for hydrolyzing PNPA increases with increase in basicity of the zinc bound hydroxides [Zn(II)±OH], yielding a linear BroÈnsted plot. Free hydroxide ®ts well on this plot. A similar plot was obtained with the enzyme mimics. The BroÈnsted relationships indicate that the Zn(II)±OH in the catalytic systems hydrolyzes the ester by direct nucleophilic attack on the ester carbonyl of cyclodextrin-bound but not Zn(II)-coordinated PNPA. # 2000 Published by Elsevier Science Ltd. All rights reserved.
Introduction Carboxypeptidase (CPA) is a much studied zinc protease which cleaves polypeptide substrates and esters. 1 The essential constituents of the active site of the enzyme are a hydrophobic pocket for substrate recognition and binding, the zinc ion that is coordinated to two imidazole nitrogens of His-69 and His-196 and the carboxylate of Glu-72, and the catalytically essential carboxylate of Glu-270.1 A water molecule is coordinated to the zinc ion as the fourth ligand. The scissile peptide carbonyl group of the incoming substrate is known to coordinate also to the zinc ion, which results in the activation of the carbonyl carbon for a nucleophilic attack. Much has been learned about CPA through kinetic and mechanistic studies, site-directed mutangenesis, and X-ray crystallography, but details of the chemistry that occurs in the catalytic process is still a matter of debate.
Recently, Kimura et al. reported that Zn(II) forms a stable and structurally well de®ned complex with [12]aneN3 and a molecule of water.3 The pKa value of the zinc bound water molecule was determined to be 7.3 (Fig. 2).4 This water molecule is thus a strong nucleophile comparable to the Zn(II) bound water molecule of carbonic anhydrase and the complex was thought to be an excellent mimic of the enzyme. Furthermore, Kimura et al. demonstrated that the Zn(II)±[12]aneN3 complex accelerates the hydrolysis of not only p-nitrophenyl acetate (PNPA) but also methyl acetate in neutral aqueous medium.4 We thought that [12]aneN3 can also mimic the three ligands that coordinate to the Zn(II) at the active site of
Christianson and Lipscomb proposed that the zinc bound water molecule that is activated by the metal ion and the carboxylate of Glu-270 serves as a nucleophile, attacking at the carbonyl carbon of the scissile peptide bond of substrate to form a tetrahedral transition state that collapses to products (Fig. 1).2
*Corresponding author. Tel.: +82-562-279-2101; fax: +82-562-2795877; e-mail: [email protected]
Figure 1. Schematic representation of the general base mechanism proposed for the enzymatic hydrolysis of peptide substrate by carboxypeptidase A.
0968-0896/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII: S0968-0896(99)00310-7
648
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
CPA and thus may be useful for the construction of CPA models. However, since CPA bears the P10 hydrophobic pocket for substrate recognition and binding as discussed above, it was thought to be necessary for the mimics of CPA to have a binding pocket that is structurally complimentary to the substrate in addition to the Zn(II) ligand. b-Cyclodextrin (b-CD) is a cyclic oligosaccharide having the overall shape of a truncated cone.5 The cone is hydrophobic and forms an inclusion complex with a phenyl ring in water. Thus, b-CD was thought to be a viable mimic of the hydrophobic pocket at the CPA active site. In this paper we describe compounds 4, 5,6 and 6 that are constructed as mimics of CPA by linking a macrocyclicpolyamine to the primary side of b-CD followed by the addition of Zn(II), and their evaluation as CPA models.
Results and Discussion Compound 1 was synthesized by allowing 6-deoxy-Otosyl-b-CD prepared by the literature method7 with a
modi®cation to react with [12]aneN3 at 100 C in DMF solution. The crude product, thus obtained, was puri®ed by CM±Sephadex chromatography followed by recrystallization from water/methanol. The structure and the purity of the product were con®rmed by 1H NMR, 13C NMR, mass spectroscopy and elemental analysis. In a similar fashion, compounds 2 and 3 were synthesized. Zinc complexes, 4, 5 and 6 were prepared in situ by adding an equimolar amount of zinc chloride to the pH 7.0 Tris buer solution containing the respective macrocyclicpolyamine, i.e. 1, 2 and 3. The enzyme mimic (catalyst) is thought to bind the substrate to form a complex that undergoes a chemical reaction to yield products with regeneration of the catalyst (Scheme 1). The breakdown of the complex to the product occurs with the ®rst-order rate constant kcat. The rate constant for the conversion of the substrate into the product in the absence of the catalyst is represented by kun. The observed ®rst order rate constant (kobs) and the dissociation constant (Km) of the complex may be expressed by eqs (1) and (2), respectively. In
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
these equations, [P], [C]0 and [S]0 represent the concentration of the product, enzyme mimic and substrate, respectively. dP=dt kobs S0 CS kun S0 kcat CS
1
Km C0 S0 =CS
2
Since CS C0 S0 =Km , eq (1) may be rewritten to give eq (3) which may be further transformed to eq (4): ÿ ÿ 3 kobs Km kun kcat C0 = Km C0 kobs ÿ kun =C0 ÿkobs =Km kcat =Km
4
The kinetic studies were carried out at 25 C in pH 7.0 buered solution using PNPA as substrate. The hydrolysis of the substrate generates p-nitrophenolate which was monitored spectrophotometrically at 400 nm. Under the conditions of a large excess of catalyst over the substrate (S0 1:0 10ÿ4 M), good pseudo-®rstorder rate constants (kobs) were obtained. The saturation kinetics were observed for all three enzyme mimics as shown in Figure 3, satisfying the Michaelis±Menten kinetics required as enzyme mimics. We plotted the kobs values for PNPA hydrolysis at varied concentrations of mimic 4 against kobs ÿ kun =CPA mimic according to eq (4) and obtained kcat and Km values of 7.7210ÿ4 sÿ1 and 3.32 mM, respectively (Fig. 4) and are collected in Table 1 along with the kinetic parameters obtained with the other two mimics. Highest rate increase was observed with 4 which accelerates the hydrolysis of PNPA by 291-fold. In principle, metal ions can accel-
649
erate the rate of hydrolysis of ester by (1) activating the carbonyl group (mechanism A), (2) activating the leaving group (mechanism B), or (3) providing a metal±hydroxide nucleophile (mechanism C).8 A combination of these activations is also possible for the hydrolysis of a given ester. In order to distinguish between the possible mechanisms that might operate for the present system, we analyzed the reactivities of the Zn(II)±cyclopolyamine complexes for hydrolyzing PNPA. The second order rate constant for the Zn(II)±hydroxide promoted hydrolysis of PNPA increases with increase in the basicity of the Zn(II)±hydroxide to generate a BroÈnsted plot with b=0.35 (Fig. 5) and we found that zinc-free hydroxide ®ts well on this plot. This observation suggests,
Figure 3. Plots of kobs versus concentrations of enzyme mimic (4, 5 and 6) in the hydrolysis of PNPA.
Figure 4. Plots of (kobs±kun)/[CPA mimic]0 versus kobs for the hydrolysis of PNPA by 4, 5 and 6. In the plot the slope of the straight line represents ÿ 1=Km and the y-intercept corresponds to kcat =Km , the second order rate constant.
Table 1. Kinetic constants for the hydrolysis of PNPA at pH 7.0 (Tris buer) and 25 C in the presence of the enzyme mimic (catalyst) Figure 2. pKa values of the water molecule bound to Zn(II)-macrocyclicpolyamide.3
Catalyst Zn(II)±[12]aneN3 Zn(II)±[12]aneN4 Zn(II)±[14]aneN4 4 5 6 a
Scheme 1.
kcat (sÿ1)
7.7210ÿ4 6.2110ÿ4 1.2810ÿ4
Km (mM)
kcat =Km (sÿ1 Mÿ1)
kcat = kun a
EM (M)
3.32 3.64 3.70
3.610ÿ3 1.810ÿ3 0.7510ÿ3 23310ÿ3 17110ÿ3 34.610ÿ3
291 234 48
0.21 0.35 0.17
The value of kun for the hydrolysis of PNPA was reported to be 2.6510ÿ6 sÿ1 (Breslow, R.; Nesnas, N. Tetrahedron Lett. 1999, 40, 3335).
650
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
strongly, that the mechanism for the hydrolysis reactions involves direct nucleophilic attack of the metal-hydroxide on the uncoordinated ester (mechanism C). The pKa value of the Zn(II) bound water molecules in 4, 5 and 6 were determined to be lower compared with the pKa of the water molecule bound to the corresponding Zn(II)± cyclicpolyamines, i.e. Zn(II)±[12]aneN3, Zn(II)±[12]aneN4 and Zn(II)±[14]ane N4, respectively by about 0.05 unit. Figure 5 also shows the BroÈnsted plot (b=0.42) constructed using the pKa values obtained with the enzyme mimics. It can be seen that the Co(III)-cyclen attached bCD that was reported by Akkaya and Czarnik9 also ®ts reasonably well on the BroÈnsted plot. Hence, we hasten to
propose that the direct nucleophilic attack of the Zn(II)hydroxide on cyclodextrin-bound but not the Zn(II)coordinated ester is also in operation for the enzyme mimic promoted hydrolysis of PNPA. Contrary to expectation, the carbonyl group of PNPA in the present models does not appear to be activated by the Zn(II). The present results are in accord with the mechanism that the zinc bound water molecule at the active site of CPA is suciently nucleophilic and thus attacks directly on the scissile bond of the substrate.2,10,11 It is worth noting that by being the substrate inserted in the hydrophobic pocket of the b-CD the rate of ester hydrolysis is enhanced by about two orders of magnitude (Fig. 5).
Figure 5. BroÈnsted plots for the hydrolysis of PNPA by enzyme mimics and Zn(II)±cyclicpolyamine complexes. The second order rate constants are corrected for Zn(II)±OH concentration. a From ref 9. b Jencks, W.P.; Gilchrist, M. J. Am. Chem. Soc. 1968, 90, 2622.
The mechanistic pathway for the hydrolysis of PNPA by the mimics at neutral pH may be represented by Figure 6. Firstly, the phenyl ring of the substrate ®ts in the hydrophobic cone of b-CD to form a complex. A nucleophilic attack by the Zn(II)-hydroxide on the ester carbonyl carbon would then ensue to generate a tetrahedral transition state which collapses to the products. Since we failed to observe catalytic turnover with the present mimics, it appears that the p-nitrophenolate that is formed remains mostly in the cone of b-CD. Analogous reaction paths are also envisioned for the PNPA hydrolysis in the presence of 5 or 6.
Figure 6. Schematic representation of the PNPA hydrolysis catalyzed by 4.
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
It is well known that catalytic groups can provide much greater rate-acceleration in intramolecular reactions than in intermolecular reactions. The eective molarity (EM) de®ned as the ratio of the rate constant for the intramolecular reaction to that of the corresponding intermolecular reaction is a parameter that re¯ects the eectiveness of a catalytic system.12 Eective molarities in the range of 105±108 M are common for intramolcular nucleophilic catalytic reactions. We were interested in ®nding out what the eective molarities would be of the enzyme mimics for hydrolyzing PNPA. The EM values of 0.21, 0.35 and 0.17 M are found for 4, 5 and 6, respectively (Table 1). Although the eective molarities are large, they are not as enormous as in a carefully designed covalent system. This is not surprising in view of the loose association between the substrate and the enzyme mimics and the ¯exible linker between the metal complex and b-CD.
Conclusion We have synthesized mimics of zinc proteases by bridging Zn(II) complexes of macrocyclicpolyamines to bCD. These enzyme mimics exhibit saturation kinetics in hydrolysis of PNPA as enzymes do, and promote the hydrolysis reaction by almost 300-fold. The eective molarities of these systems range from 0.17 to 0.35 M. From the analysis of the BroÈnsted plots obtained with the mimics and the Zn(II)-macrocyclicpolyamines for the PNPA hydrolysis, it was concluded that the Zn(II)bound water molecule in the mimics attacks on the ester carbonyl of the PNPA that is CD-bound but not Zn(II)coordinated.
Experimental Melting points (MP) were determined on a Thomas± Hoover capillary MP apparatus and are uncorrected. 1 H NMR and 13C NMR spectra were recorded on a Bruker FT-NMR spectrometer (300 or 500 MHz) and chemical shifts are expressed in ppm relative to tetramethylsilane. Samples were dissolved in a mixture of deuterium oxide and deuteriomethylsulfoxide. Low resolution mass spectra were obtained with a KRATOS MS 25 RFA spectrometer. Kinetic study was carried out using a Perkin±Elmer HP 8453 UV±vis spectrometer. Elemental analyses were performed at POSTECH (CBM), Pohang, South Korea. Puri®cation of cyclodextrin derivatives were performed by CM-25 Sephadex chromatography. All chemicals were of reagent grade obtained from Aldrich Chemical Co. N,N-Dimethylformamide was distilled over magnesium sulfate and stored under nitrogen. Mono-6-deoxy-6-(p-toluenesulfonyl)- -cyclodextrin ( CD - 6 - OTs). A solution of p-toluenesulfonyl chloride (1.5 g, 7.9 mmol) in anhydrous pyridine (10 mL) was added with stirring to a solution of b-CD (5.0 g, 4.4 mmol, dried under vacuum at 100 C for 12 h) dissolved in anhydrous pyridine (50 mL). After 5 h at 4 C, pyridine was removed under reduced pressure at 40 C
651
and a light yellowish syrup was obtained. Upon the addition of acetone (300 mL) to the residue a colorless solid precipitated which was collected by suction ®ltration and washed with acetone. Recrystallization of the crude product from water yielded an analytically pure sample (1.8 g, 30% yield) as a white crystal: mp 168±170 C (dec.); 1H NMR (DMSO-d6) d 2.43 (3H, s), 3.33±3.57 (42H, m), 4.32 (6H, s), 4.84 (7H, s), 5.67 (14H, s), 7.43 (2H, d), 7.76 (2H, d); 13C NMR (DMSO-d6) d 21.80, 59.76, 71.90, 72.88, 81.36, 101.73, 127.35, 129.66, 132.47, 144.56. Mono-6-deoxy-6-(1,5,9-triazacyclododecanyl)- -cyclodextrin (1). Dried b-CD-6-OTs, (0.65 g, 0.5 mmol) was dissolved in DMF (20 mL) and 1,5,9-triazacyclododecane (0.43 g, 2.5 mmol) was added to this solution. The resulting mixture was heated at 100 C for 5 h under N2 atmosphere. After completion of the reaction, the mixture was evaporated to dryness in vacuo at 40 C and acetone (100 mL) was added to the residue. The solid thus formed was collected by suction ®ltration and washed with acetone. The residue was puri®ed by CM-25 Sephadex chromatography eluting with water followed by 0.1 M NH4OH to give an analytically pure sample (0.46 g, 72% yield) as a white solid mp 232 C (dec.); 1H NMR (D2O) d 2.08±2.18 (6H, m), 3.27±3.34 (12H, m), 3.43±4.20 (42H, m), 5.06 (7H, s); 13C NMR (D2O) d 27.48, 47.93, 59.76, 71.90, 72.88, 81.36, 101.73; Fab-MS (m/z) 1288 (M+1); anal. calcd for C51H89N3O34. 4.5H2O: C, 44.73; H, 7.21; N, 3.07. Found: C, 44.41; H, 6.98; N, 3.12. Mono-6-deoxy-6-(1,4,7,10-tetraazacyclododecanyl)- -cyclodextrin (2). This compound was synthesized in 64% yield by the same procedure as described above mp 246 C (dec.); 1H NMR (D2O) d 2.73±2.78 (16H, m), 3.45±4.18 (42H, m), 5.07 (7H, s); 13C NMR (D2O) d 46.03, 59.66, 71.88, 73.08, 81.38, 101.79; Fab-MS (m/z) 1289 (M+1); anal. calcd. for C50H88N4O34.6.5H2O: C, 42.70; H, 7.24; N, 3.98. Found: C, 42.59; H, 7.06; N, 3.99. Mono-6-deoxy-6-(1,4,8,11-tetraazacyclotetradecayl)- cyclodextrin (3). This compound was synthesized in 52% yield by the same procedure as described above: mp 254 C (dec.); 1H NMR (D2O) d 1.78±1.84 (4H, m), 2.75±2.80 (8H, m), 2.83±2.90 (8H, m), 3.45±4.20 (42H, m), 5.09 (7H, s); 13C NMR (D2O) d 29.85, 40.63, 42.82, 60.06, 70.98, 73.08, 81.86, 101.51; Fab-MS (m/z) 1317 (M+1); anal. calcd. for C52H92N4O34 6H2O: C, 43.00; H, 7.22; N, 3.86. Found : C, 42.87; H, 7.07; N, 3.92. Kinetics Hydrolysis rate of p-nitrophenyl acetate in aqueous solution in the presence of an enzymic mimic was measured by following the increase in the 400 nm absorption using a computer-linked UV spectrometer. The reaction solution was maintained at 25 C using a buffered solution containing 0.05 M Tris buer (pH 7.0), the ionic strength of which was adjusted to 0.10 with sodium chloride. The typical procedure was as follows: the enzyme mimic, zinc chloride (both ®nal concentrations of 1.0±10.0 mM) and p-nitrophenyl acetate (the ®nal concentration of 0.1 mM) in 5% acetonitrile
652
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
solution were mixed and the absorption at 400 nm was recorded as a function of time. The observed rate constants (kobs) were obtained directly from the computer interfaced UV spectrometer. All experiments were run in duplicate and the tabulated data represent the average of these experiments. Acknowledgements This work was supported by the Pohang University of Science and Technology and the Korea Science and Engineering Foundation. The authors express their thanks to Professor Jik Chin for the valuable help in the preparation of this manuscript. References and Notes 1. Guiocho, F. A.; Lipscomb, W. N. Adv. Protein Chem. 1971, 25, 1.
2. Christianson, D. W.; Lipscomb, W. N. Acc. Chem. Res. 1989, 22, 62. 3. Kimura, E.; Koike, T. Comments Inorg. Chem. 1991, 11, 285. 4. Kimura, E.; Shiota, T.; Koike, T.; Shiro, M.; Kodama, M. J. Am. Chem. Soc. 1990, 112, 5805. 5. Bender, M. L.; Komiyama, M. Cyclodextrin Chemistry; Springer-Verlag: Berlin, 1978: pp 34±41. 6. This compound has been reported in the literature (Rosenthal, M. I.; Czarnik, A. W. J. Inclusion. Phenom. 1991, 10, 119. 7. Matsui, Y.; Okimoto, A. Bull. Chem. Soc. Jpn. 1978, 51, 3030. 8. Chin, J. Acc. Chem. Res. 1991, 24, 145. 9. Akkaya, E. U.; Czarnick, A. W. J. Am. Chem. Soc. 1988, 110, 8553. 10. Breslow, R.; Wernick, D. L. J. Am. Chem. Soc. 1976, 98, 259. 11. Auld, D. S.; Galdes, A.; Geoghegan, K. F.; Holmquist, B.; Martinelli, R. A.; Ballee, B. L. Proc. Natl Acad. Sci. USA 1984, 81, 5041. 12. Ables, R. H.; Frey, P. A.; Jencks, W. P. Biochemistry; Jones and Bartlett Publisher: Boston, 1992, pp 135±137.
Origin of Rate-Acceleration in Ester Hydrolysis with Metalloprotease Mimics Dong H. Kim* and Soo Suk Lee Center for Biofunctional Molecules and Department of Chemistry, Pohang University of Science and Technology, Pohang 790-784, South Korea Received 23 August 1999; accepted 26 November 1999
AbstractÐMimics of carboxypeptidase A, a prototypical metalloprotease, were synthesized by linking macrocyclicpolyamines to the primary side of b-cyclodextrin followed by complexing with Zn(II). These enzyme mimics exhibit saturation kinetics in hydrolysis of p-nitrophenyl acetate (PNPA) and enhance the rate of hydrolysis reaction by almost 300-fold. The eective molarities (EM) of the mimics range from 0.2 to 1.9 M. Origin of the rate acceleration was examined: the reactivity of Zn(II) complexes of [12]aneN3 [12]aneN4, and [14]aneN4 for hydrolyzing PNPA increases with increase in basicity of the zinc bound hydroxides [Zn(II)±OH], yielding a linear BroÈnsted plot. Free hydroxide ®ts well on this plot. A similar plot was obtained with the enzyme mimics. The BroÈnsted relationships indicate that the Zn(II)±OH in the catalytic systems hydrolyzes the ester by direct nucleophilic attack on the ester carbonyl of cyclodextrin-bound but not Zn(II)-coordinated PNPA. # 2000 Published by Elsevier Science Ltd. All rights reserved.
Introduction Carboxypeptidase (CPA) is a much studied zinc protease which cleaves polypeptide substrates and esters. 1 The essential constituents of the active site of the enzyme are a hydrophobic pocket for substrate recognition and binding, the zinc ion that is coordinated to two imidazole nitrogens of His-69 and His-196 and the carboxylate of Glu-72, and the catalytically essential carboxylate of Glu-270.1 A water molecule is coordinated to the zinc ion as the fourth ligand. The scissile peptide carbonyl group of the incoming substrate is known to coordinate also to the zinc ion, which results in the activation of the carbonyl carbon for a nucleophilic attack. Much has been learned about CPA through kinetic and mechanistic studies, site-directed mutangenesis, and X-ray crystallography, but details of the chemistry that occurs in the catalytic process is still a matter of debate.
Recently, Kimura et al. reported that Zn(II) forms a stable and structurally well de®ned complex with [12]aneN3 and a molecule of water.3 The pKa value of the zinc bound water molecule was determined to be 7.3 (Fig. 2).4 This water molecule is thus a strong nucleophile comparable to the Zn(II) bound water molecule of carbonic anhydrase and the complex was thought to be an excellent mimic of the enzyme. Furthermore, Kimura et al. demonstrated that the Zn(II)±[12]aneN3 complex accelerates the hydrolysis of not only p-nitrophenyl acetate (PNPA) but also methyl acetate in neutral aqueous medium.4 We thought that [12]aneN3 can also mimic the three ligands that coordinate to the Zn(II) at the active site of
Christianson and Lipscomb proposed that the zinc bound water molecule that is activated by the metal ion and the carboxylate of Glu-270 serves as a nucleophile, attacking at the carbonyl carbon of the scissile peptide bond of substrate to form a tetrahedral transition state that collapses to products (Fig. 1).2
*Corresponding author. Tel.: +82-562-279-2101; fax: +82-562-2795877; e-mail: [email protected]
Figure 1. Schematic representation of the general base mechanism proposed for the enzymatic hydrolysis of peptide substrate by carboxypeptidase A.
0968-0896/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII: S0968-0896(99)00310-7
648
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
CPA and thus may be useful for the construction of CPA models. However, since CPA bears the P10 hydrophobic pocket for substrate recognition and binding as discussed above, it was thought to be necessary for the mimics of CPA to have a binding pocket that is structurally complimentary to the substrate in addition to the Zn(II) ligand. b-Cyclodextrin (b-CD) is a cyclic oligosaccharide having the overall shape of a truncated cone.5 The cone is hydrophobic and forms an inclusion complex with a phenyl ring in water. Thus, b-CD was thought to be a viable mimic of the hydrophobic pocket at the CPA active site. In this paper we describe compounds 4, 5,6 and 6 that are constructed as mimics of CPA by linking a macrocyclicpolyamine to the primary side of b-CD followed by the addition of Zn(II), and their evaluation as CPA models.
Results and Discussion Compound 1 was synthesized by allowing 6-deoxy-Otosyl-b-CD prepared by the literature method7 with a
modi®cation to react with [12]aneN3 at 100 C in DMF solution. The crude product, thus obtained, was puri®ed by CM±Sephadex chromatography followed by recrystallization from water/methanol. The structure and the purity of the product were con®rmed by 1H NMR, 13C NMR, mass spectroscopy and elemental analysis. In a similar fashion, compounds 2 and 3 were synthesized. Zinc complexes, 4, 5 and 6 were prepared in situ by adding an equimolar amount of zinc chloride to the pH 7.0 Tris buer solution containing the respective macrocyclicpolyamine, i.e. 1, 2 and 3. The enzyme mimic (catalyst) is thought to bind the substrate to form a complex that undergoes a chemical reaction to yield products with regeneration of the catalyst (Scheme 1). The breakdown of the complex to the product occurs with the ®rst-order rate constant kcat. The rate constant for the conversion of the substrate into the product in the absence of the catalyst is represented by kun. The observed ®rst order rate constant (kobs) and the dissociation constant (Km) of the complex may be expressed by eqs (1) and (2), respectively. In
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
these equations, [P], [C]0 and [S]0 represent the concentration of the product, enzyme mimic and substrate, respectively. dP=dt kobs S0 CS kun S0 kcat CS
1
Km C0 S0 =CS
2
Since CS C0 S0 =Km , eq (1) may be rewritten to give eq (3) which may be further transformed to eq (4): ÿ ÿ 3 kobs Km kun kcat C0 = Km C0 kobs ÿ kun =C0 ÿkobs =Km kcat =Km
4
The kinetic studies were carried out at 25 C in pH 7.0 buered solution using PNPA as substrate. The hydrolysis of the substrate generates p-nitrophenolate which was monitored spectrophotometrically at 400 nm. Under the conditions of a large excess of catalyst over the substrate (S0 1:0 10ÿ4 M), good pseudo-®rstorder rate constants (kobs) were obtained. The saturation kinetics were observed for all three enzyme mimics as shown in Figure 3, satisfying the Michaelis±Menten kinetics required as enzyme mimics. We plotted the kobs values for PNPA hydrolysis at varied concentrations of mimic 4 against kobs ÿ kun =CPA mimic according to eq (4) and obtained kcat and Km values of 7.7210ÿ4 sÿ1 and 3.32 mM, respectively (Fig. 4) and are collected in Table 1 along with the kinetic parameters obtained with the other two mimics. Highest rate increase was observed with 4 which accelerates the hydrolysis of PNPA by 291-fold. In principle, metal ions can accel-
649
erate the rate of hydrolysis of ester by (1) activating the carbonyl group (mechanism A), (2) activating the leaving group (mechanism B), or (3) providing a metal±hydroxide nucleophile (mechanism C).8 A combination of these activations is also possible for the hydrolysis of a given ester. In order to distinguish between the possible mechanisms that might operate for the present system, we analyzed the reactivities of the Zn(II)±cyclopolyamine complexes for hydrolyzing PNPA. The second order rate constant for the Zn(II)±hydroxide promoted hydrolysis of PNPA increases with increase in the basicity of the Zn(II)±hydroxide to generate a BroÈnsted plot with b=0.35 (Fig. 5) and we found that zinc-free hydroxide ®ts well on this plot. This observation suggests,
Figure 3. Plots of kobs versus concentrations of enzyme mimic (4, 5 and 6) in the hydrolysis of PNPA.
Figure 4. Plots of (kobs±kun)/[CPA mimic]0 versus kobs for the hydrolysis of PNPA by 4, 5 and 6. In the plot the slope of the straight line represents ÿ 1=Km and the y-intercept corresponds to kcat =Km , the second order rate constant.
Table 1. Kinetic constants for the hydrolysis of PNPA at pH 7.0 (Tris buer) and 25 C in the presence of the enzyme mimic (catalyst) Figure 2. pKa values of the water molecule bound to Zn(II)-macrocyclicpolyamide.3
Catalyst Zn(II)±[12]aneN3 Zn(II)±[12]aneN4 Zn(II)±[14]aneN4 4 5 6 a
Scheme 1.
kcat (sÿ1)
7.7210ÿ4 6.2110ÿ4 1.2810ÿ4
Km (mM)
kcat =Km (sÿ1 Mÿ1)
kcat = kun a
EM (M)
3.32 3.64 3.70
3.610ÿ3 1.810ÿ3 0.7510ÿ3 23310ÿ3 17110ÿ3 34.610ÿ3
291 234 48
0.21 0.35 0.17
The value of kun for the hydrolysis of PNPA was reported to be 2.6510ÿ6 sÿ1 (Breslow, R.; Nesnas, N. Tetrahedron Lett. 1999, 40, 3335).
650
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
strongly, that the mechanism for the hydrolysis reactions involves direct nucleophilic attack of the metal-hydroxide on the uncoordinated ester (mechanism C). The pKa value of the Zn(II) bound water molecules in 4, 5 and 6 were determined to be lower compared with the pKa of the water molecule bound to the corresponding Zn(II)± cyclicpolyamines, i.e. Zn(II)±[12]aneN3, Zn(II)±[12]aneN4 and Zn(II)±[14]ane N4, respectively by about 0.05 unit. Figure 5 also shows the BroÈnsted plot (b=0.42) constructed using the pKa values obtained with the enzyme mimics. It can be seen that the Co(III)-cyclen attached bCD that was reported by Akkaya and Czarnik9 also ®ts reasonably well on the BroÈnsted plot. Hence, we hasten to
propose that the direct nucleophilic attack of the Zn(II)hydroxide on cyclodextrin-bound but not the Zn(II)coordinated ester is also in operation for the enzyme mimic promoted hydrolysis of PNPA. Contrary to expectation, the carbonyl group of PNPA in the present models does not appear to be activated by the Zn(II). The present results are in accord with the mechanism that the zinc bound water molecule at the active site of CPA is suciently nucleophilic and thus attacks directly on the scissile bond of the substrate.2,10,11 It is worth noting that by being the substrate inserted in the hydrophobic pocket of the b-CD the rate of ester hydrolysis is enhanced by about two orders of magnitude (Fig. 5).
Figure 5. BroÈnsted plots for the hydrolysis of PNPA by enzyme mimics and Zn(II)±cyclicpolyamine complexes. The second order rate constants are corrected for Zn(II)±OH concentration. a From ref 9. b Jencks, W.P.; Gilchrist, M. J. Am. Chem. Soc. 1968, 90, 2622.
The mechanistic pathway for the hydrolysis of PNPA by the mimics at neutral pH may be represented by Figure 6. Firstly, the phenyl ring of the substrate ®ts in the hydrophobic cone of b-CD to form a complex. A nucleophilic attack by the Zn(II)-hydroxide on the ester carbonyl carbon would then ensue to generate a tetrahedral transition state which collapses to the products. Since we failed to observe catalytic turnover with the present mimics, it appears that the p-nitrophenolate that is formed remains mostly in the cone of b-CD. Analogous reaction paths are also envisioned for the PNPA hydrolysis in the presence of 5 or 6.
Figure 6. Schematic representation of the PNPA hydrolysis catalyzed by 4.
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
It is well known that catalytic groups can provide much greater rate-acceleration in intramolecular reactions than in intermolecular reactions. The eective molarity (EM) de®ned as the ratio of the rate constant for the intramolecular reaction to that of the corresponding intermolecular reaction is a parameter that re¯ects the eectiveness of a catalytic system.12 Eective molarities in the range of 105±108 M are common for intramolcular nucleophilic catalytic reactions. We were interested in ®nding out what the eective molarities would be of the enzyme mimics for hydrolyzing PNPA. The EM values of 0.21, 0.35 and 0.17 M are found for 4, 5 and 6, respectively (Table 1). Although the eective molarities are large, they are not as enormous as in a carefully designed covalent system. This is not surprising in view of the loose association between the substrate and the enzyme mimics and the ¯exible linker between the metal complex and b-CD.
Conclusion We have synthesized mimics of zinc proteases by bridging Zn(II) complexes of macrocyclicpolyamines to bCD. These enzyme mimics exhibit saturation kinetics in hydrolysis of PNPA as enzymes do, and promote the hydrolysis reaction by almost 300-fold. The eective molarities of these systems range from 0.17 to 0.35 M. From the analysis of the BroÈnsted plots obtained with the mimics and the Zn(II)-macrocyclicpolyamines for the PNPA hydrolysis, it was concluded that the Zn(II)bound water molecule in the mimics attacks on the ester carbonyl of the PNPA that is CD-bound but not Zn(II)coordinated.
Experimental Melting points (MP) were determined on a Thomas± Hoover capillary MP apparatus and are uncorrected. 1 H NMR and 13C NMR spectra were recorded on a Bruker FT-NMR spectrometer (300 or 500 MHz) and chemical shifts are expressed in ppm relative to tetramethylsilane. Samples were dissolved in a mixture of deuterium oxide and deuteriomethylsulfoxide. Low resolution mass spectra were obtained with a KRATOS MS 25 RFA spectrometer. Kinetic study was carried out using a Perkin±Elmer HP 8453 UV±vis spectrometer. Elemental analyses were performed at POSTECH (CBM), Pohang, South Korea. Puri®cation of cyclodextrin derivatives were performed by CM-25 Sephadex chromatography. All chemicals were of reagent grade obtained from Aldrich Chemical Co. N,N-Dimethylformamide was distilled over magnesium sulfate and stored under nitrogen. Mono-6-deoxy-6-(p-toluenesulfonyl)- -cyclodextrin ( CD - 6 - OTs). A solution of p-toluenesulfonyl chloride (1.5 g, 7.9 mmol) in anhydrous pyridine (10 mL) was added with stirring to a solution of b-CD (5.0 g, 4.4 mmol, dried under vacuum at 100 C for 12 h) dissolved in anhydrous pyridine (50 mL). After 5 h at 4 C, pyridine was removed under reduced pressure at 40 C
651
and a light yellowish syrup was obtained. Upon the addition of acetone (300 mL) to the residue a colorless solid precipitated which was collected by suction ®ltration and washed with acetone. Recrystallization of the crude product from water yielded an analytically pure sample (1.8 g, 30% yield) as a white crystal: mp 168±170 C (dec.); 1H NMR (DMSO-d6) d 2.43 (3H, s), 3.33±3.57 (42H, m), 4.32 (6H, s), 4.84 (7H, s), 5.67 (14H, s), 7.43 (2H, d), 7.76 (2H, d); 13C NMR (DMSO-d6) d 21.80, 59.76, 71.90, 72.88, 81.36, 101.73, 127.35, 129.66, 132.47, 144.56. Mono-6-deoxy-6-(1,5,9-triazacyclododecanyl)- -cyclodextrin (1). Dried b-CD-6-OTs, (0.65 g, 0.5 mmol) was dissolved in DMF (20 mL) and 1,5,9-triazacyclododecane (0.43 g, 2.5 mmol) was added to this solution. The resulting mixture was heated at 100 C for 5 h under N2 atmosphere. After completion of the reaction, the mixture was evaporated to dryness in vacuo at 40 C and acetone (100 mL) was added to the residue. The solid thus formed was collected by suction ®ltration and washed with acetone. The residue was puri®ed by CM-25 Sephadex chromatography eluting with water followed by 0.1 M NH4OH to give an analytically pure sample (0.46 g, 72% yield) as a white solid mp 232 C (dec.); 1H NMR (D2O) d 2.08±2.18 (6H, m), 3.27±3.34 (12H, m), 3.43±4.20 (42H, m), 5.06 (7H, s); 13C NMR (D2O) d 27.48, 47.93, 59.76, 71.90, 72.88, 81.36, 101.73; Fab-MS (m/z) 1288 (M+1); anal. calcd for C51H89N3O34. 4.5H2O: C, 44.73; H, 7.21; N, 3.07. Found: C, 44.41; H, 6.98; N, 3.12. Mono-6-deoxy-6-(1,4,7,10-tetraazacyclododecanyl)- -cyclodextrin (2). This compound was synthesized in 64% yield by the same procedure as described above mp 246 C (dec.); 1H NMR (D2O) d 2.73±2.78 (16H, m), 3.45±4.18 (42H, m), 5.07 (7H, s); 13C NMR (D2O) d 46.03, 59.66, 71.88, 73.08, 81.38, 101.79; Fab-MS (m/z) 1289 (M+1); anal. calcd. for C50H88N4O34.6.5H2O: C, 42.70; H, 7.24; N, 3.98. Found: C, 42.59; H, 7.06; N, 3.99. Mono-6-deoxy-6-(1,4,8,11-tetraazacyclotetradecayl)- cyclodextrin (3). This compound was synthesized in 52% yield by the same procedure as described above: mp 254 C (dec.); 1H NMR (D2O) d 1.78±1.84 (4H, m), 2.75±2.80 (8H, m), 2.83±2.90 (8H, m), 3.45±4.20 (42H, m), 5.09 (7H, s); 13C NMR (D2O) d 29.85, 40.63, 42.82, 60.06, 70.98, 73.08, 81.86, 101.51; Fab-MS (m/z) 1317 (M+1); anal. calcd. for C52H92N4O34 6H2O: C, 43.00; H, 7.22; N, 3.86. Found : C, 42.87; H, 7.07; N, 3.92. Kinetics Hydrolysis rate of p-nitrophenyl acetate in aqueous solution in the presence of an enzymic mimic was measured by following the increase in the 400 nm absorption using a computer-linked UV spectrometer. The reaction solution was maintained at 25 C using a buffered solution containing 0.05 M Tris buer (pH 7.0), the ionic strength of which was adjusted to 0.10 with sodium chloride. The typical procedure was as follows: the enzyme mimic, zinc chloride (both ®nal concentrations of 1.0±10.0 mM) and p-nitrophenyl acetate (the ®nal concentration of 0.1 mM) in 5% acetonitrile
652
D. H. Kim, S. S. Lee / Bioorg. Med. Chem. 8 (2000) 647±652
solution were mixed and the absorption at 400 nm was recorded as a function of time. The observed rate constants (kobs) were obtained directly from the computer interfaced UV spectrometer. All experiments were run in duplicate and the tabulated data represent the average of these experiments. Acknowledgements This work was supported by the Pohang University of Science and Technology and the Korea Science and Engineering Foundation. The authors express their thanks to Professor Jik Chin for the valuable help in the preparation of this manuscript. References and Notes 1. Guiocho, F. A.; Lipscomb, W. N. Adv. Protein Chem. 1971, 25, 1.
2. Christianson, D. W.; Lipscomb, W. N. Acc. Chem. Res. 1989, 22, 62. 3. Kimura, E.; Koike, T. Comments Inorg. Chem. 1991, 11, 285. 4. Kimura, E.; Shiota, T.; Koike, T.; Shiro, M.; Kodama, M. J. Am. Chem. Soc. 1990, 112, 5805. 5. Bender, M. L.; Komiyama, M. Cyclodextrin Chemistry; Springer-Verlag: Berlin, 1978: pp 34±41. 6. This compound has been reported in the literature (Rosenthal, M. I.; Czarnik, A. W. J. Inclusion. Phenom. 1991, 10, 119. 7. Matsui, Y.; Okimoto, A. Bull. Chem. Soc. Jpn. 1978, 51, 3030. 8. Chin, J. Acc. Chem. Res. 1991, 24, 145. 9. Akkaya, E. U.; Czarnick, A. W. J. Am. Chem. Soc. 1988, 110, 8553. 10. Breslow, R.; Wernick, D. L. J. Am. Chem. Soc. 1976, 98, 259. 11. Auld, D. S.; Galdes, A.; Geoghegan, K. F.; Holmquist, B.; Martinelli, R. A.; Ballee, B. L. Proc. Natl Acad. Sci. USA 1984, 81, 5041. 12. Ables, R. H.; Frey, P. A.; Jencks, W. P. Biochemistry; Jones and Bartlett Publisher: Boston, 1992, pp 135±137.